Portfolio Selection and Optimization Based on Factor Investing Strategy

Abstract

This thesis's aim is to explore the practical application of factor investment strategies in portfolio construction for individual investors. The traditional portfolio construction method based on historical values is becoming increasingly inadequate in coping with the nowadays complex investment market. Factor investing, an emerging investment concept, aims to capture the performance of underlying fundamental, technical, and systematic risk factors to optimize the portfolio effectively. This thesis discusses the construction of an investment analysis process suitable for individual investors, explaining stock market returns by means of various factors and discriminating factor characteristics through machine learning. It draws on the latest research reports from investment banks on multi-factor model testing. It utilizes quantitative platforms such as Ricequant and Joinquant to maintain the universality and usability of the research environment.